PR Wrangler: Kristi

  • We’re having issues where PRs are included commits from the destination/master branch. Jason sent email about re-making pull requests when commits show up in both the source and destination branches. For example this PR contains 112 commits. If one then looks at this one it currently shows 13 commits.
  • Sahil, Kyle and Kristi want to use readthedocs.org for our documentation.
    • The team would like to convert all of the documentation to Markdown instead of rst, since it’s easier to read/parse. Jason discovered that if we convert Readme.rst to Readme.md, we need an additional converter for the install process.
  • There’s been some discussion on the oldest Issue, https://github.com/CCI-MOC/haas/issues/39.
  • For console: current bug is that /var/run/haas_console_logs is not being created and made writable by “haas_user”. Jason sent out email on this and filed an issue on github. Sahil will make a PR for this.
  • Need a way of dividing up the dev haas resources. Everyone should be running their own haas dev server on top of the development haas master (moc-haas-master.bu.edu), ie - using “haas serve” and not under apache/wsgi. Discussed text file vs wiki for nodes and VLANs.
    • We’re going to use an MOC wiki page to divide resources. Protocol will be to reserve it on the wiki and commit first, then use it.
    • Jason will update with details
  • Kyle’s network ACL’s PR (#554) is ready for review. Reviewers are George + (Ian or Jason)
  • Sahil’s client library PR is going to be closed, and he is going to start submitting the client library in shorter chunks. This week will just be query calls.
  • Need additional discussion for VPN solution regarding first implementation (VMs vs processes).
    • Ron and Orran will come up with a couple of times and send them out.
  • Naming contest: need to solidify on a new name soon. Jason will send polls this week.
  • Jason is working on fixing the args verification problem.
  • We are looking forward Kristi’s work (hopefully ready by mid next week) to integrate apache into our continuous integration, which should prevent problems like the above from reappearing.